Cyclopentadiene (CPD) is currently a minor byproduct of liquid fed steam cracking (for example, naphtha and heavier feed). As existing and new steam cracking facilities shift to lighter feeds, less CPD is produced while demand for CPD is rising. High cost due to supply limitations impacts the potential end product use of CPD in polymers. More CPD-based polymer product could be produced, if additional CPD could be produced, at unconstrained rates and preferably at a cost lower than recovery from steam cracking. Cyclopentane and cyclopentene also have high value as solvents while cyclopentene may be used as a co-monomer to produce polymers and as a starting material for other high value chemicals.
It would be advantageous to be able to produce cyclic C5 compounds including CPD as the primary product from plentiful C5 feedstock using a catalyst system to produce CPD while minimizing production of light (C4−) byproducts. While lower hydrogen content feedstock (for example, cyclics, alkenes, dialkenes) could be preferred because the reaction endotherm is reduced and thermodynamic constraints on conversion are improved, non-saturates are more expensive than saturate feedstock. Linear C5 skeletal structure is preferred over branched C5 skeletal structures due to both reaction chemistry and the lower value of linear C5 relative to branched C5 (due to octane differences). An abundance of C5 is available from unconventional gas and shale oil as well as reduced use in motor fuels due to stringent emissions requirements. C5 feedstock may also be derived from bio-feeds.
Dehydrogenation technologies are currently used to produce mono-olefins and di-olefins from C3 and C4 alkanes, but not cyclic mono-olefins or cyclic di-olefins. A typical process uses Pt/Sn supported on alumina as the active catalyst. Another useful process uses chromia on alumina. See, B. V. Vora, “Development of Dehydrogenation Catalysts and Processes”, Topics in Catalysis, vol. 55, pp. 1297-1308, 2012; and J. C. Bricker, “Advanced Catalytic Dehydrogenation Technologies for Production of Olefins”, Topics in Catalysis, vol. 55, pp. 1309-1314, 2012.
Still another common process uses Pt/Sn supported on Zn and/or Ca aluminate to dehydrogenate propane. While these processes are successful in dehydrogenating alkanes, they do not perform cyclization which is critical to producing CPD. Pt—Sn/alumina and Pt—Sn/aluminate catalysts exhibit moderate conversion of n-pentane, but such catalyst have poor selectivity and yield to cyclic C5 products.
Pt supported on chlorided alumina catalysts are used to reform low octane naphtha to aromatics such as benzene and toluene. See, U.S. Pat. No. 3,953,368 (Sinfelt), “Polymetallic Cluster Compositions Useful as Hydrocarbon Conversion Catalysts.” While these catalysts are effective in dehydrogenating and cyclizing C6 and higher alkanes to form C6 aromatic rings, they are less effective in converting acyclic C5S to cyclic C5S. These Pt on chlorided alumina catalysts exhibit low yields of cyclic C5 and exhibit deactivation within the first two hours of time on stream. Cyclization of C6 and C7 alkanes is aided by the formation of an aromatic ring, which does not occur in C5 cyclization. This effect may be due in part to the much higher heat of formation for CPD, a cyclic C5, as compared to benzene, a cyclic C6, and toluene, a cyclic C7. This is also exhibited by Pt/Ir and Pt/Sn supported on chlorided alumina. Although these alumina catalysts perform both dehydrogenation and cyclization of C6+ species to form C6 aromatic rings, a different catalyst will be needed to convert acyclic C5 to cyclic C5.
Ga-containing ZSM-5 catalysts are used in a process to produce aromatics from light paraffins. A study by Kanazirev et al. showed n-pentane is readily converted over Ga2O3/H-ZSM-5. See Kanazirev et al., “Conversion of C8 aromatics and n-pentane over Ga2O3/H-ZSM-5 mechanically mixed catalysts,” Catalysis Letters, vol. 9, pp. 35-42, 1991. No production of cyclic C5 was reported while upwards of 6 wt % aromatics were produced at 440° C. and 1.8 hr−1 WHSV. Mo/ZSM-5 catalysts have also been shown to dehydrogenate and/or cyclize paraffins, especially methane. See, Y. Xu, S. Liu, X. Guo, L. Wang, and M. Xie, “Methane activation without using oxidants over Mo/HZSM-5 zeolite catalysts,” Catalysis Letters, vol. 30, pp. 135-149, 1994. High conversion of n-pentane using Mo/ZSM-5 was demonstrated with no production of cyclic C5 and high yield to cracking products. This shows that ZSM-5-based catalysts can convert paraffins to a C6 ring, but not necessarily to produce a C5 ring.
U.S. Pat. No. 5,254,787 (Dessau) introduced the NU-87 catalyst used in the dehydrogenation of paraffins. This catalyst was shown to dehydrogenate C2-5 and C6+ to produce their unsaturated analogs. A distinction between C2-5 and C6+ alkanes was made explicit in this patent: dehydrogenation of C2-5 alkanes produced linear or branched mono-olefins or di-olefins whereas dehydrogenation of C6+ alkanes yielded aromatics. U.S. Pat. No. 5,192,728 (Dessau) involves similar chemistry, but with a tin-containing crystalline microporous material. As with the NU-87 catalyst, C5 dehydrogenation was only shown to produce linear or branched, mono-olefins or di-olefins and not CPD.
U.S. Pat. No. 5,284,986 (Dessau) introduced a dual-stage process for the production of cyclopentane and cyclopentene from n-pentane. An example was conducted wherein the first stage involved dehydrogenation and dehydrocyclization of n-pentane to a mix of paraffins, mono-olefins and di-olefins, and naphthenes over a Pt/Sn-ZSM-5 catalyst. This mixture was then introduced to a second-stage reactor consisting of Pd/Sn-ZSM-5 catalyst where dienes, especially CPD, were converted to olefins and saturates. Cyclopentene was the desired product in this process, whereas CPD was an unwanted byproduct.
U.S. Pat. No. 2,438,398; U.S. Pat. No. 2,438,399; U.S. Pat. No. 2,438,400; U.S. Pat. No. 2,438,401; U.S. Pat. No. 2,438,402; U.S. Pat. No. 2,438,403, and U.S. Pat. No. 2,438,404 (Kennedy) disclosed production of CPD from 1,3-pentadiene over various catalysts. Low operating pressures, low per pass conversion, and low selectivity make this process undesirable. Additionally, 1,3-pentadiene is not a readily available feedstock, unlike n-pentane. See also, Kennedy et al., “Formation of Cyclopentadiene from 1,3-Pentadiene,” Industrial & Engineering Chemistry, vol. 42, pp. 547-552, 1950.
Fel'dblyum et al. in “Cyclization and dehydrocyclization of C5 hydrocarbons over platinum nanocatalysts and in the presence of hydrogen sulfide,” Doklady Chemistry, vol. 424, pp. 27-30, 2009, reported production of CPD from 1,3-pentadiene, n-pentene, and n-pentane. Yields to CPD were as high as 53%, 35%, and 21% for the conversion of 1,3-pentadiene, n-pentene, and n-pentane respectively at 600° C. on 2% Pt/SiO2. While initial production of CPD was observed, drastic catalyst deactivation within the first minutes of the reaction was observed. Experiments conducted on Pt-containing silica show moderate conversion of n-pentane over Pt—Sn/SiO2, but with poor selectivity and yield to cyclic C5 products. The use of H2S as a 1,3-pentadiene cyclization promoter was presented by Fel'dblyum, infra, as well as in Marcinkowski, “Isomerization and Dehydrogenation of 1,3-Pentadiene,” M.S., University of Central Florida, 1977. Marcinkowski showed 80% conversion of 1,3-pentadiene with 80% selectivity to CPD with H2S at 700° C. High temperature, limited feedstock, and potential of products containing sulfur that would later need scrubbing make this process undesirable.
López et al. in “n-Pentane Hydroisomerization on Pt Containing HZSM-5, HBEA and SAPO-11,” Catalysis Letters, vol. 122, pp. 267-273, 2008, studied reactions of n-pentane on Pt-containing zeolites including H-ZSM-5. At intermediate temperatures (250-400° C.), they reported efficient hydroisomerization of n-pentane on the Pt-zeolites with no discussion of cyclopentenes formation. It is desirable to avoid this deleterious chemistry as branched C5 do not produce cyclic C5 as efficiently as linear C5, as discussed above.
Li et al. in “Catalytic dehydroisomerization of n-alkanes to isoalkenes,”Journal of Catalysis, vol. 255, pp. 134-137, 2008, also studied n-pentane dehydrogenation on Pt-containing zeolites in which Al had been isomorphically substituted with Fe. These Pt/[Fe]ZSM-5 catalysts were efficient dehydrogenating and isomerizing n-pentane, but under the reaction conditions used, no cyclic C5 were produced and undesirable skeletal isomerization occurred.
In view of this state of the art, there remains a need for a process to convert acyclic C5 feedstock to non-aromatic, cyclic C5 hydrocarbon, namely CPD, preferably at commercial rates and conditions. Further, there is a need for a catalytic process targeted for the production of cyclopentadiene which generates cyclopentadiene in high yield from plentiful C5 feedstocks without excessive production of C4− cracked products and with acceptable catalyst aging properties. This invention meeting this and other needs.